Posted January 3, 1996
------------------------

     Q METHODOLOGY AS THE FOUNDATION FOR A SCIENCE OF SUBJECTIVITY


                            Steven R. Brown

                         Kent State University



+----------------------------------------------------------------------+
|                                                                      |
|     Abstract:  William Stephenson's 1935 letter to the Editor of     |
|  Nature, published 60 years ago, contains within its four short      |
|  paragraphs all the essentials for a science of subjectivity.  Fo-   |
|  cusing on two studies -- on the controversies surrounding animal    |
|  experimentation and of problem selection in policy analysis -- il-  |
|  lustrations are provided of the new phenomena brought to light      |
|  through "inverted" factor analysis, and of the advantages of ex-    |
|  perimentation which Q methodology enjoys.  A proposed study on      |
|  food habits demonstrates how experimental probes can be made into   |
|  the cognitive and orectic aspects of nutrition and food choice.     |
|  The conclusion is reached that Q methodology remains the founda-    |
|  tion of the study of subjective behavior.                           |
|                                                                      |
+----------------------------------------------------------------------+
|                                                                      |
|  Read at the Eleventh International Conference of the International  |
|  Society for the Scientific Study of Subjectivity, College of Medi-  |
|  cine, University of Illinois, Chicago, 12-14 October 1995.          |
|                                                                      |
+----------------------------------------------------------------------+



                        Q Methodology, 1935-1995


It has now been 60 years since William Stephenson took pen in hand and
informed the editor of Nature of that bright idea of his that eventually
came to be known as Q methodology (see Figure 1).  It has been an
eventful 60 years -- World War II, the Korean War, the Vietnam War; the
administrations of 11 U.S. presidents -- Roosevelt, Truman, Eisenhower,
Kennedy, Johnson, Nixon, Ford, Carter, Reagan, Bush, and Clinton; space
flight, cures for polio and other diseases, and significant advances
against others; and developments in information technology of all kinds
(television, satellites, computers) with consequent developments in the
technical sophistication of mass publics.  In my graduate days, we used
to leave off data which the computer center staff would then key-punch
into cards and analyze for us; we would pick up our box of cards and
output a few days later.  Today we talk directly to the computer and get
our output in seconds.

    All of this is progress in the objective mode -- of hardware and
software, cures for cancer, moon walks, and warfare -- but we would be
hard-pressed to point to much in the way of advances in understanding
subjectivity during this same period, i.e., of humankind's understanding
of itself in its own terms.  Academic psychology has become preoccupied
with cognition and brain physiology, requiring high levels of expert
knowledge; psychoanalysis has been chased from the psychology depart-
ments and medical schools, or has otherwise been incorporated by depart-
ments of English, which have often rendered it esoteric and
unintelligible; and postmodernism has likewise removed understanding of
human thought beyond the bounds to which ordinary individuals can as-
pire.  As a consequence, the bookshelves in the psychology and philoso-


+--------------------------------------------------+
|                                                  |
|           Technique of Factor Analysis           |
|                                                  |
|     Factor  analysis  is  a  subject upon which  |
|  Prof. G. H. Thomson, Dr. Wm. Brown and  others  |
|  have  frequently  written  letters  to Nature.  |
|  This analysis is  concerned  with  a  selected  |
|  population  of  n individuals each of whom has  |
|  been measured in  m  tests.    The  (m)(m-1)/2  |
|  intercorrelations  for  these  m variables are  |
|  subjected to either a Spearman or other factor  |
|  analysis.                                       |
|     The technique, however,  can  also  be  in-  |
|  verted.   We begin with a population of n dif-  |
|  ferent tests (or essays, pictures,  traits  or  |
|  other  measurable  material), each of which is  |
|  measured or scaled  by  m  individuals.    The  |
|  (m)(m-1)/2  intercorrelations are then factor-  |
|  ised in the usual way.                          |
|     This inversion  has  interesting  practical  |
|  applications.   It brings the factor technique  |
|  from group and field work into the laboratory,  |
|  and reaches into spheres of work hitherto  un-  |
|  touched  or not amenable to factorisation.  It  |
|  is especially valuable  in  experimental  aes-  |
|  thetics and in educational psychology, no less  |
|  than in pure psychology.                        |
|     It   allows  a  completely  new  series  of  |
|  studies to be made on  the  Spearman  'central  |
|  intellective  factor'  (g),  and  also  allows  |
|  tests to be made of the Two Factor Theorem un-  |
|  der greatly improved experimental  conditions.  |
|  Data  on these and other points are to be pub-  |
|  lished in due course in the British Journal of  |
|  Psychology.                                     |
|                                                  |
|                                W. Stephenson     |
|                                                  |
| Psychological Laboratory,                        |
|    University College,                           |
|       Gower Street,                              |
|       London, W.C.1.                             |
|          June 28.                                |
|                                                  |
+--------------------------------------------------+
   From Nature, 24 August 1935, vol. 136, p. 297

Fig. 1.  Stephenson's initial statement.


phy sections of bookstores are increasingly stocked with mysticism,
spiritualism, and the occult.  The situation has all the earmarks of
Marxian alienation:  we produce our own ordinary thoughts, but then seem
to lack the thinking capacity necessary to repurchase those same
thoughts when they are returned to us in the manufactured form of philo-
sophical or psychological theories about thought.

    Anniversaries provide occasions to reflect on beginnings and to take
stock, and were we to do that with Stephenson's letter, we would see
that much of the conceptual and technical equipment necessary to move
forward is there, explicitly or implicitly, in that short note.  His in-
novation required a few illustrations which subsequent publications pro-
vided, but these were primarily elaborations of the logic already
tightly packed into that 1935 letter.

    We recognize the first paragraph of Stephenson's letter as a summary
of conventional factor analysis (R methodology), against which he wished
to contrast another method, and there is no disputing that the conven-
tional version has remained dominant in the intervening years, invari-
ably to the exclusion of the alternative (Stephenson, 1990a, 1990b):
The prevailing use of factor analysis remains the factorization of those
(m)(m-1)/2 intercorrelations which have arisen from the measurement of
individuals on tests.


                          The Second Paragraph

However, the second paragraph opens the door to something truly new, de-
spite the efforts of some critics to restrict its impact by associating
it with the reciprocal of R factor analysis.  From the outset, as we
see, Stephenson equated Q methodology (which did not yet have a name)
with materials as measured by individuals:  It was the m individuals who
were doing the measuring rather than being measured, and so subjectivity
was implicated from the very beginning.  And since the (m)(m-1)/2 inter-
correlations were then to be "factorised in the usual way," the result-
ing factor composites, like the individual measures which comprised
them, were also drenched with subjectivity.

    Take for illustrative purposes the hot topic of animal experimenta-
tion and the views about it which have been expressed during the past
100 years and more (as made conveniently available in Magel, 1989):

    For the scientist, animals are only organisms concealing prob-
      lems which the scientist must solve. (Claude Bernard,
      physiologist, 1865)

    The cry of an injured lark would stab me to the heart, but when
      we probe the mysteries of life and acquire new truth, the sov-
      ereignty of the end in view carries all before it. (Louis
      Pasteur [1822-1895], chemist)

    Anesthetics should, of course, be used, but experiments must be
      allowed to go forward due to the benefits to mankind. (Charles
      Darwin, naturalist, 1876)

    I could not kill or hurt any living creature needlessly, nor de-
      stroy any beautiful thing. (John Ruskin, professor of litera-
      ture, 1885, who resigned his position at Oxford the day after
      vivisection was introduced)

    The pain which vivisection inflicts upon unconsenting animals is
      sufficient justification for my enmity without looking fur-
      ther. (Mark Twain, author, 1901)

    The rights of the helpless, even though they be brutes, must be
      protected by those who have superior power. (William James,
      psychologist, 1909)

And so on up to today's battles over animal rights, which have become so
heated as to threaten the unity of the animal rights movement itself,
dividing it (or so it is asserted) into those supporting more humane
treatment and those taking a more absolutist stance (see, for example,
Jasper and Nelkin, 1992).

    Stephenson's assertion that conventional factor analysis "can also
be inverted" constituted a metaphor which was wasted on various of his
critics, some of whom subsequently took satisfaction in pointing out
that the matrix manipulation wasn't really an inversion but a transposi-
tion.  It was, of course, neither; rather, it was a conceptual shift
that moved the focus of attention from "variables" of a transindividual
character (e.g., intelligence or volition in some abstract sense) to how
concrete individuals were actually thinking and reacting to the materi-
als placed in front of them -- enjoying or approving of some, disagree-
ing with others, and feeling neither here nor there about still others.
Thoughts and ideas have feeling tones, as Stout (1899, pp. 562-580) long
ago noted, and it was these thoughts interlaced with feeling that
Stephenson's "inversion" brought into prominence.

    These feelings of approval and disapproval, of pleasure and unpleas-
ure, are at the heart of subjectivity, and they are concrete and imme-
diately experienceable rather than transcendental.  In Psychology Down
the Ages, Spearman (1937, p. 449) referred to pleasure and unpleasure as
perhaps the sole examples of sensory states that have achieved patency
"down the ages," and one of the elegant features of Q methodology is
that it provides a formal model of pleasure/unpleasure -- in the form of
the Q sort.  In fact, it is hard to conceive of any more elegant way to
represent subjectivity -- hence the title of this paper:  "Q Methodology
as the Foundation for a Science of Subjectivity."

    What do we find when, as Stephenson instructed, the (m)(m-1)/2 cor-
relations among persons are "factorised in the usual way"?  A Q sample
drawn from Magel's (1989) collection was administered to members of an
animal rights group, plus acquaintances of theirs who they thought might
hold different views (Brown & Squire, in progress).  It is not espe-
cially surprising that one of the factors consists of the so-called
Absolutists, i.e., members of the animal rights vanguard who believe
that animals have absolute rights, and who exceed the other two factors
in their support of the following views:

       Animal experimentation usually produces trivial results, uses
    intrinsically objectionable means, and morally desensitizes
    teachers and students....  The pain which vivisection inflicts
    upon unconsenting animals is reason enough to oppose it, without
    looking farther....  Extrapolating results from animals to hu-
    mans can be inaccurate and misleading -- physiologically,
    psychologically, and in other ways....  Animal experimenters de-
    fend their practices because they have professional reputations
    to protect.  How can they admit their error when their status
    rests on a lifetime of publications with their names on them?

The Absolutists are naturally concerned with pain inflicted on animals,
and so eagerly embrace Mark Twain's conclusion that the pain involved is
reason enough to oppose vivisection.  The enemy in all this is obviously
science -- more specifically, scientists -- and the ideological stance
of this factor was so strong that it ended up taking precedence over the
following two statements, much to the chagrin of persons comprising this
factor once the results were shown to them (scores for factors I, II,
and III, respectively):

    I could not kill or hurt any living creature needlessly. ( 0  +3
    +4)

    There must be policies to provide adequate protection for lab
    animals against unnecessary abuse. ( 0  +4  +4)

It is understandable that some animal rightists might downplay the sec-
ond statement (so as not to lend legitimacy to animal labs), but it came
as a surprise that this group did not embrace among its first principles
Ruskin's declaration against the needless hurting of living things.

    Only two people defined the second factor, but one was a researcher
who had spent a career devoted to science and the observation of simian
behavior in far-away places, and a pro-science stance is the most visi-
ble characteristic of this factor, as shown in statements receiving sig-
nificantly higher scores:

    Anesthetics should be used whenever possible....  Animals will
    be used rationally if both science and ethics answer the same
    question: Is this animal the best experimental system for the
    job?....  Where possible, use should be made of humans them-
    selves as experimental subjects in lieu of animals....  Much of
    the opposition to animal experimentation can be traced to a gen-
    eral lack of knowledge about science generally, and about what
    actually goes on in an experiment specifically.

At the negative end of the Science factor were denials that the external
grant system props up trivial research, that faculty rarely address the
ethics of experimentation, and that experimentation morally desensitizes
teachers and students alike.  But behind the scientific bravado were
signs of uncertainty, as in factor II's zero score to the following
statement:  "There are ethical problems in animal experimentation due to
the continuity of human and other animal life" (+3  0 +3).

    Finally, factor III: the Humanists -- in the sense that they elevate
the welfare of Homo sapiens above that of other creatures (scores for
factors I to III, respectively):

    The justification of animal experiments is in the expected bene-
    fits to humans. (-1  -1  +2)

    There is more sacredness, surely, about one human being than
    about all the other animal species put together. (-4  -4  +1)

    Where possible, use should be made of humans themselves as ex-
    perimental subjects in lieu of animals. (+1  +3  -4)

The first two statements receive relatively low scores, but relatively
high in contrast to I and II; however, the factor's strong negative re-
action to the last statement suggests a reticence to make too strong a
proclamation about human superiority.  Elsewhere in the statement array,
the Humanists displayed a marked tendency to side first with the
Absolutists (e.g., in questioning vivisection and the legitimacy of sci-
ence) and then with the Scientists (on issues associated with policing
against unnecessary abuse).  The Humanists wish not to harm animals un-
less absolutely necessary for human welfare.

    It is worth noting that there were visible points of consensus to-
ward which the three factors gravitated for their divergent motivations
-- that we need to find alternatives to animal experimentation, that
boycotting companies relying on animal research is a legitimate consumer
strategy, and that universities have an obligation to promote alterna-
tive methods (e.g., to dissection) so that students need not be forced
to harm animals.  The strong scientific assertion -- that animal re-
search is ethically mandatory -- was roundly condemned by all three fac-
tors.

    As is apparent, therefore, the inversion to which Stephenson pointed
in the second paragraph of his 1935 letter brings new features of human
existence to light, and those features are more common and understand-
able.  Rather than abstract traits and faculties, such as intelligence
and volition, we are brought face to face with such familiar phenomena
as animal rightists so certain in their criticisms of others that they
forget kindness as their own first principle, of scientists boldly de-
fending their practices (but underneath perhaps feeling more vulner-
able), and ambivalent humanists wishing to defend animals against pain
yet feeling conflicted when animal rights come into conflict with human
welfare.  These are not faculties or abstract social categories, but re-
cognizable ways of living in the world.  In introducing this inversion,
therefore -- in which individuals are now doing the measuring, rather
than being measured -- Stephenson set factor analysis and the study of
human behavior spinning on a new axis.


                          The Third Paragraph

The third paragraph in Stephenson's letter promises new and practical
applications in regions of inquiry not previously examined by factor
procedures, with aesthetics and educational psychology especially sin-
gled out, and he notes that his inversion "brings the factor technique
from group and field work into the laboratory."  This serves as a re-
minder of Stephenson's training as a physicist, and of the skills and
reasoning of the experimentalist which training of this kind brings with
it.  The work of previous factor analysts, in their search for primary
abilities and dimensions, necessarily involved large masses of data.  As
Thomas (1935) wrote, in the same year as Stephenson's letter, "The only
method which promises any finality of decision is to get together as
many different people as possible, to test them for as many different
abilities as possible, to compare the results, and to base one's views
of ability upon the outcome" (p. 25).  This takes time, however, and
therefore stands as an impediment to a truly experimental science, as
Stephenson (1935b) pointed out:

       A research student may spend two years isolating a single
    factor....  One cannot perform an experiment today and use its
    results for another tomorrow.  Tests cannot be changed in the
    way that a chemist changes the mixtures in his test-tube to try
    out his hypotheses.  In short, the present-day technique lacks
    the pliability that the energetic experimentalist wants at his
    command.  It is a device for massive field work, and not for the
    clinic, the laboratory, or for rapid and subtle experimenta-
    tion.... (p. 18)

Stephenson then wondered whether there was an alternative:  "Can we make
factor studies on a few individuals, and bring the methods of corre-
lations and factor analysis into the laboratory?" (p. 18).  His answer
to his own question was, of course, in the affirmative, and this affir-
mation was backed up with Q methodology.

    A recent inquiry demonstrates how results can be obtained in a
timely fashion, as required for "rapid and subtle experimentation," and
without recourse to the ponderous volume of observations characteristic
of field work.  In this instance, members of a senior seminar in policy
science were invited to nominate possible policy problems for more de-
tailed study.  Following a round-robin nomination process, which netted
27 possible topics, the 23 seminar members were then instructed to Q
sort the topics in terms of the interest each had for them.  Four fac-
tors (two of them bipolar) were ready for discussion within 48 hours.

    The factors represent patterns of interest which the students have
developed and which guide their choice of those problem areas in which
to invest effort for the remainder of the semester.  The character of
one of the bipolar factors emerges clearly in the issues dominating the
two ends of the factor array:

    Positive pole:  impact of drug traffic on Latin American socie-
      ties, policy toward drugs, problems of illegal immigration,
      gun control, changes in the intelligence community

    Negative pole:  full employment, impact of societal change on
      women's employment, crackdown on "deadbeat dads," poverty and
      hunger, animal rights

This factor pits the status quo against those who wish to alter it.  The
young policy scientists defining the positive end of the factor are anx-
ious to study those social problems (drugs, guns, illegal immigration)
that pose a threat to U.S. society and also those institutions (e.g.,
the CIA) which defend it:  Social pathologies such as drugs and illegal
immigration are seen as signs of societal decay.  Students at the oppo-
site end of the factor wish to study those problems (unemployment, pov-
erty) which they see as consequences of the status quo, and they
identify with those social categories (women, abandoned families, ani-
mals) which are disadvantaged under current arrangements.  For these
participants (unlike those at the other end of the factor), poverty and
homelessness are causes, and drugs, guns, and illegal immigrants are
consequences, rather than the other way around.

    As noted previously, ideas have feeling tone, and this applies no
less to policy interests and to the operant factors indicative of inter-
est schemata that hold the individual policy problems in place.  (As
Stephenson µ1986º observed in this regard, "...concern is with states-
of-feeling in Q, not with specific feeling attached normatively to par-
ticular statements" [p. 537].)  Along with their policy Q sorts, the
seminar members were also to provide biographical sketches so as to il-
luminate those features of their career lines that might reveal the
genesis of their policy interests, and these personal essays made clear
in most cases how the blooms of policy interest initially took root and
were nourished in the fertile soil of significant life experiences.  For
example:

       Respondent 16 (factor A-positive):  This person wished to
    study the Central Intelligence Agency, in which he had been in-
    terested since he was young.  He had been strongly influenced by
    his father, whom he considered his mentor and friend, and who
    had been a military and civilian police officer as well as asso-
    ciate of personnel in the World War II Office of Strategic Ser-
    vices (predecessor to the CIA).  He has applied for a job with
    the CIA, in which he hopes to enjoy a career.

       Respondent 20 (factor A-negative):  This person begins her
    essay with the declaration that "I am a woman who comes from a
    family of very strong and independent women."  That she is pre-
    pared for "all the harrassment I will endure being a woman in a
    powerful job" she attributes to her mother.  Her parents are di-
    vorced, and although her father was "always there for me when I
    needed him," she has nonetheless developed an interest in
    "deadbeat fathers."  Her favorite high school teacher, a woman
    and animal rights activist, extended her indignation concerning
    the powerless to include animals.  She concludes that "one day I
    want to be powerful and change people's views about what women
    can and cannot do in this world."

    There were other factors that revealed sentiments for emotional vs.
intellectual issues, for international vs. domestic policies, and for
alternatives to policies and institutions that were malfunctioning
(e.g., social security and welfare), and in virtually all cases the fac-
tors were a function of demonstrable interests which were in turn embed-
ded, like hands in gloves, to the particulars of individuals' lives.

    Such is the nature of apperception, i.e., the "readiness to perceive
this or that in relation to prior systems of interest" (Stephenson,
1967, p. 149).  Introduced by Leibnitz, apperception was incorporated
into educational theory by Herbart in the early 19th century, and was a
central principle in the systematic psychologies of James Ward (1920,
pp. 308-312) and especially G.F. Stout (1896, pp. 110-167), but until
Stephenson introduced Q methodology this important concept was without
operational foundation.  It is, of course, not restricted to education:
As Stephenson (1967) points out, newsreading is awash with
apperceptiveness, with readers turning to the editorial, sports, fash-
ion, and gossip sections as their interests direct.  Contemporary inno-
vations such as electronic Gopher sites and the World Wide Web merely
expand the apperceptive arena.

    But what about the requirements for "rapid and subtle experimenta-
tion" which Stephenson promised in his 1935 letter?  In a worked example
(Brown, 1989), a policy analyst was invited to Q sort a set of organiza-
tional problems under a variety of conditions of performance:  Which
problems interest you most?  Which would the authorities regard as most
sensitive?  Which are most amenable to solution? are most urgent? are
most apt to get worse? are most upsetting? would contribute most to hu-
man dignity? and so forth.  Each of these experimental performances --
with one or two completed each day for a week -- places the Q sorter in
a different hypothetical situation, thereby altering the experiment com-
parable to the way in which "a chemist changes the mixtures in his test-
tube to try out his hypotheses" (Stephenson, 1935b, p. 18).  Although
these experimental possibilities were announced 60 years ago, Q's poten-
tial in this regard has hardly been explored.


                          The Fourth Paragraph

Spearman's g referred to a general ability that entered into everything
from writing a play to tying a shoelace.  It was heavily cognitive (as
opposed to orectic) and constituted the major prong of the "two-
factor theorem"; the other prong was s, the family of specific factors,
i.e., specific to each ability:  Some people can add numbers quickly,
for instance, which implicates g but also an ability specific to adding
but that does not carry over to a facility for remembering names, which
in turn has its own s (Thomas, 1935, p. 28).

    I wish to suggest a study (rather than report one) that involves
general intelligence (not in an IQ sense) as well as specific abilities,
and both cognitive and orectic capabilities.  The topic is food, which
has physiologic, psychologic, and cultural features.

    Food is used in symbolic and socially-expressive ways (Douglas,
1984; Fieldhouse, 1986), and we might begin for comparative purposes
with two housewives from different social classes, so as to provide room
for contrasts to emerge due to differences in specific knowledge and
abilities.  Q technique has been employed in food and nutrition studies,
but not in especially effective ways.  Sutnick (1981), for example, con-
structed a Q sample composed of the names of 25 foods which 266 exper-
imental students (plus 246 controls) ranked by preference and then again
for nutritive value; however, Sutnick then summed across all exper-
imental and all control subjects for purposes of comparison, thereby
preventing operant categories from emerging.  Simpson (1989) utilized
pictures of food rather than food names, but also averaged across arbi-
trary categories (mothers of low vs. normal weight/length children).
Fitzgerald (1977) introduced a Q-sort Food Choice Game with pictures of
food, and also recommended multiple conditions of instruction; however,
he did not report his findings.  None of these really approximates the
"greatly improved experimental conditions" which Stephenson's method
made possible.

    For purposes of structuring a Q sample, pictures of foods or food
names could be used, the former illustrated by Fitzgerald and Simpson
(supra), the latter by Sutnick and also Meiselman and Waterman (1978).
How, next, to measure?  Most efforts have naturally been categorical,
involving such concepts as flavor, prestige, and convenience (Krondl &
Lau, 1982), or satiety, price, and beliefs about health (Reaburn, Krondl
& Lau, 1979) and the like.  Schutz, Rucker and Russell (1975) are in the
proper spirit when they suggest "allowing respondents themselves to gen-
erate the classifications" (p. 52), but ultimately they take averages
across their 200 respondents before proceeding to an R factor analysis.

    However, Schutz, Rucker and Russell, before taking their averages,
do suggest a number of conditions under which food might be consumed,
and these could provide the basis for genuine experimentation:

       Possible conditions of experiment (cognitive):  When I want
    something nutritious....  When I want something inexpensive....
    When I want something spicy....  When I want something easy to
    prepare... and so forth.  (All of these imply special know-
    ledge.)

       Possible conditions of experiment (orectic):  When I'm trying
    to please my husband ("a way to a man's heart")....  For the
    kids when they're being rambunctious....  For the kids when
    they've been especially good....  Putting on a good show for
    special guests....  When I'm depressed....  For snacking in
    front of the TV....  Things I sometimes have a craving for...
    etc.  (These conditions implicate orectic strivings and de-
    sires.)

    On several occasions, Stephenson noted his intellectual indebtedness
to J.R. Kantor (e.g., Stephenson, 1984), and it is not difficult to im-
agine that he would have been sympathetic to efforts to include Kantor's
ideas along with those of Spearman.  Kantor (1959, p. 16), it will be
recalled, formalized a psychological event as PE = C (k, sf, rf, hi, st,
md), and it is easy to conceive of conditions of instruction (using the
above food Q sample) that would cover Kantor's PE dimensions:

       Which of the foods do you most enjoy preparing (sf)?  Which
    make your mouth water just thinking about them (rf)?  Which did
    you like most as a child, or were most apt to have when you and
    your husband were first married (hi)?  Which would you be apt to
    serve during the work week, or for Sunday brunch, or with candle
    light, or on a picnic, or when you are dining out, etc. (st)?
    Which fill the kitchen with their aroma (md)?

And on and on.  It would be a rather straightforward matter, first to
factor the Spearman Q sorts (based on central intellective and orectic
principles), and then the Kantor Q sorts (based on PE considerations),
and then to refactor the combined arrays from these two solutions to see
for ourselves precisely in what ways the two converge.

    It is also worth considering that the subjectivities involved in Q
sorts produced under conditions of instruction such as the above can be
supplemented by certain "objectivities."  Nutritionists, for instance,
could be enlisted to provide Q sorts representing caloric, fat, nutri-
tional and other contents.  Parents often give in to demanding children
by giving them high-sugar foods that spiral into even more energetic de-
mands, and depressed and emotionally abused women often turn to foods
which exacerbate their situations; anorexics and bulimics likewise have
eating patterns which, under the magnifying glass of factor analysis,
might be shown to have structures capable of modification.

    Lewin (1943), in his classic study of food habits, was concerned to
determine the conditions under which bad eating patterns might be al-
tered, and in this regard, experiments of the kind implied in the propo-
sitional sets above would bring us closer to understanding the social
constraints to be overcome.  Lewin noted at the outset that one of the
chief difficulties was "to find categories which have meaning from the
nutritionist's point of view and still are in line with the everyday
terms in which the housewife thinks and acts" (p. 36).  Lewin's solution
was to interview a relatively large number of individuals and to take
averages over social categories, such as Czech, low-income White, and
Negro groups.  Whereas studies of that kind have their place, they can
profitably be supplemented with experimental studies designed to test
two-factor, interbehavioral, or any other theorems that suggest promis-
ing leads.


                           Concluding Remarks

And the same holds true for any other problem area in the entire domain
of human endeavor.  William Stephenson's innovation, announced in 1935,
was designed to provide an instrumental basis for studies not heretofore
covered by objective procedures.  It was not a simple matrix transposi-
tion, as so many since have contended and continue to believe:  It was
totally new, and it was so precisely because it focused a penetrating
beam of light on subjectivity -- on individuals measuring rather than
being measured.

    This conceptual inversion for the first time placed us in a position
to carry out studies of considerable sophistication and within shortened
time frames that continue to induce wonder even after more than a half
century of technical and conceptual development in the human sciences.
And the range of topics amenable to Q is just as impressive -- cogni-
tive, conative, volitional (to use the older language) right up to mod-
ern and postmodern applications in narrative and discourse analysis,
hermeneutics, media hegemony, literary criticism, decision making, and
all else, including phenomena which fall under the rubric of quantum
consciousness and mind (but which really amount to subjective
communicability).

    There is no other method or theory which matches Q's versatility or
reach, and which comports so well with the principles and concepts of
contemporary science (including anti-science!), and so it remains today,
as it was envisioned 60 years ago -- the foundation for the scientific
study of subjectivity.



Footnotes
------------------------------------------------------------------------

1.  Stephenson frequently cited James Ward as a source of inspiration,
    and G.F. Stout was Ward's successor at Cambridge University.  His
    chapter on "Feeling-Tone of Ideas" notes that thought can "occasion
    changes in the common sensibility" either by reviving past feelings
    (as in reminiscence) or as a result of the thought process itself
    (e.g., by solving a problem).  Feeling tone can also be painful, as
    when the thinker is conflicted or confused.  We would add to Stout's
    inventory the pleasure derived through mere repetition, as when an
    ideologue uses the Q sort as an opportunity to assert a predigested
    doctrine once again.

2.  Cognition involves perceiving and thinking, affection involves feel-
    ing, conation striving, and volition willing, the latter three com-
    bined into the category orexis (Thomas, 1935, pp. 7-13).



References
------------------------------------------------------------------------

Brown, S.R. (1989).  A feeling for the organism: Understanding and in-
  terpreting political subjectivity.  Operant Subjectivity, 12,  81-97.
Brown, S.R. & Squire, T. (in progress).  Dimensions of the animal rights
  movement.
Douglas, M. (1984).  Standard social uses of food: Introduction.  In M.
  Douglas (Ed.), Food in the social order (pp. 1-39).  New York: Russell
  Sage Foundation.
Fieldhouse, P. (1986).  Food & nutrition: Customs & culture.  London:
  Croom Helm.
Fitzgerald, T.K. (1977).  Anthropological approaches to the study of
  food habits: Some methodological issues.  In T.K. Fitzgerald (Ed.),
  Nutrition and anthropology in action (pp. 69-78).  Assen, The
  Netherlands: Van Gorcum.
Jasper, J.M. & Nelkin, D. (1992).  The animal rights crusade: The growth
  of a moral protest.  New York: Free Press.
Kantor, J.R. (1959).  Interbehavioral psychology (2nd ed.).  Granville,
  OH: Principia Press.
Krondl, M. & Lau, D. (1982).  Social determinants in human food se-
  lection.  In L.M. Barker (Ed.), The psychobiology of human food se-
  lection (pp. 139-151).  Westport, CT: AVI.
Lewin, K. (1943).  Forces behind food habits and methods of change.  In
  National Research Council, The problem of changing food habits (pp.
  35-65).  Washington, DC: National Research Council, National Academy
  of Sciences.
Magel, C.R. (1989).  Keyguide to information sources in animal rights.
  London: Mansell.
Meiselman, H.L. & Waterman, D. (1978).  Food preferences of enlisted
  personnel in the Armed Forces.  Journal of the American Dietetic Asso-
  ciation, 73, 621-629.
Reaburn, J.A., Krondl, M. & Lau, D. (1979).  Social determinants in food
  selection.  Journal of the American Dietetic Association, 74, 637-641.
Schutz, H.G., Rucker, M.H. & Russell, G.F. (1975).  Food and food-use
  classification systems.  Food Technology, 29, 50-64.
Simpson, S.H. (1989).  Use of Q-sort methodology in cross-cultural nu-
  trition and health research.  Nursing Research, 38, 289-290.
Spearman, C. (1937).  Psychology down the ages (Vol. 1).  London:
  Macmillan.
Stephenson, W. (1935a).  Technique of factor analysis.  Nature, 136,
  297.
Stephenson, W. (1935b).  Correlating persons instead of tests.  Charac-
  ter and Personality, 4, 17-24.
Stephenson, W. (1967).  The play theory of mass communication.  Chicago:
  University of Chicago Press.
Stephenson, W. (1984).  Methodology for statements of problems: Kantor
  and Spearman conjoined.  Psychological Record, 34, 575-588.
Stephenson, W. (1986).  William James, Niels Bohr, and complementarity:
  II -- Pragmatics of a thought.  Psychological Record, 36, 529-543.
Stephenson, W. (1990a).  Fifty years of exclusionary psychometrics: I. Q
  technique.  Operant Subjectivity, 13, 105-120.
Stephenson, W. (1990b).  Fifty years of exclusionary psychometrics: II.
  Developments.  Operant Subjectivity, 13 141-162.
Stout, G.F. (1896).  Analytic psychology (vol. 2).  London: Swan
  Sonnenschein.
Stout, G.F. (1899).  A manual of psychology.  London: University Corre-
  spondence College Press.
Sutnick, M.R. (1981).  The Q sort technique applied to nutrition atti-
  tudes investigation.  Journal of Nutrition Education, 13(suppl.),
  S102-S105.
Thomas, F.C. (1935).  Ability and knowledge: The standpoint of the
  London School.  London: Macmillan.
Ward, J. (1920).  Psychological principles (2nd ed.).  Cambridge:
  Cambridge University Press.